A method and system for beam tracking based on spatial positioning of a drone

Abstract

The application provides a kind of based on unmanned aerial vehicle space positioning's beam tracking method and system, wherein the method includes constructing unmanned aerial vehicle ground millimeter wave communication system model and receiving spectral efficiency model;Coordinate system is constructed to establish the relationship between unmanned aerial vehicle attitude and transmission beam azimuth and elevation angle;Unmanned aerial vehicle attitude is estimated by extended Kalman filtering and back propagation estimation;Unmanned aerial vehicle trajectory is predicted;According to the trajectory prediction result of unmanned aerial vehicle, the extended Kalman filtering estimation result and the back propagation estimation result of unmanned aerial vehicle attitude are input into receiving spectral efficiency model, and the real spectral efficiency is obtained, to quantify the beam tracking effect with real spectral efficiency.The application estimates the attitude of unmanned aerial vehicle by combining back propagation neural network and extended Kalman filtering algorithm, and predicts the trajectory of unmanned aerial vehicle at base station, combines the estimation and prediction results, obtains spatial beam angle through coordinate conversion, forms analog beam forming vector, and realizes accurate beam tracking.

Description

Technical Field

[0001] This invention belongs to the field of UAV beam tracking technology, and particularly relates to a beam tracking method and system based on UAV spatial positioning. Background Technology

[0002] With the rapid development of drones, their tasks have become more diversified, placing higher demands on their data throughput. Compared to other communication technologies, millimeter-wave wireless communication technology offers advantages such as high bandwidth, high-speed transmission, and ease of integration, which can meet the communication needs of drones. However, the significant advantage of millimeter-wave communication stems from its high directionality, achieved through large-scale array antenna beamforming technology to form directional beams and overcome high-frequency path loss. However, the narrow beam of millimeter-wave communication is susceptible to the effects of rapid drone movement, making real-time alignment difficult. Therefore, real-time beam tracking technology is needed to reduce link interruptions and maintain reliable links.

[0003] Currently, millimeter-wave beam tracking methods are mainly divided into three categories: The first category is Bayesian statistical beam tracking methods. These methods can predict the current position based on previous position and velocity when tracking highly mobile users, thus achieving beam tracking. The second category is information-assisted beam tracking methods. These methods utilize auxiliary information obtained from sensors to reconfigure directional antennas and switch beam directions, reducing the beam search space and avoiding frequent link interruptions. The third category is machine learning-based beam tracking methods. These methods have lower complexity and can capture channel dynamics, making them well-adaptable to different environments. These methods typically track the beam arrival / departure angle directly. However, in the highly dynamic environment of UAV communication, directly tracking the spatial beam arrival / departure angle is difficult because it is not only related to the UAV's position but also closely related to its attitude. Errors in UAV attitude estimation can significantly impact beam tracking. Furthermore, since the UAV's position cannot be obtained in real time at the base station, its position needs to be predicted. Summary of the Invention

[0004] This invention addresses the shortcomings of existing technologies by providing a beam tracking method and system based on UAV spatial positioning.

[0005] In a first aspect, the present invention provides a beam tracking method based on UAV spatial positioning, comprising:

[0006] Construct a model of a UAV-to-ground millimeter-wave communication system and a model of its receiving spectrum efficiency;

[0007] Construct a coordinate system and establish the relationship between the UAV's attitude and the azimuth and elevation angles of the transmitted beam;

[0008] The azimuth and elevation angles of the UAV's transmitted beam are determined by extended Kalman filter estimation and backpropagation estimation of the UAV's attitude.

[0009] The trajectory of the drone is predicted to obtain the drone trajectory prediction result;

[0010] The UAV trajectory prediction results, UAV attitude extended Kalman filter estimation results, and backpropagation estimation results are input into the receiver spectral efficiency model to obtain the true spectral efficiency, and the beam tracking effect is quantified by the true spectral efficiency.

[0011] Furthermore, the construction of the UAV-to-ground millimeter-wave communication system model and the receiving spectrum efficiency model includes:

[0012] Constructing the transmit antenna array response model:

[0013]

[0014] Among them, the drone's transmitting antenna is equipped with M t ×N t A uniform planar antenna array with individual antenna elements; This is the response vector of the transmitting antenna array; θ is the azimuth angle of the transmitted beam at the UAV location. t The elevation angle of the transmitted beam at the UAV location; the array element spacing of the transmitting antenna d1 = λ c / 2;λ c The carrier wavelength is 1 ≤ m t ≤M t ; 1≤n t ≤N t ;m t n is the number of rows in the uniform planar antenna array for transmitting antennas; t The number of columns in the uniform planar antenna array for transmitting antennas;

[0015] Construct a receive antenna array response model:

[0016]

[0017] Among them, the base station receiving antenna is equipped with M r ×N r A uniform planar antenna array with individual antenna elements; This is the response vector of the receiving antenna array; θ is the azimuth angle of the receiving beam at the base station. r The elevation angle of the receiving beam at the base station; the array element spacing of the receiving antenna d2 = λ c / 2;1≤m r ≤M r ; 1≤n r ≤Nr ;m r n is the number of rows in the uniform planar antenna array for receiving antennas; r The number of columns in the uniform planar antenna array for receiving antennas;

[0018] Based on the transmit antenna array response model and the receive antenna array response model, a time-varying channel matrix for UAV-to-ground millimeter-wave communication is constructed:

[0019]

[0020] Among them, H k Let be the time-varying channel matrix for UAV-to-ground millimeter-wave communication at time k; gain k D is the complex path gain at time k; k Let be the transmission distance between the drone and the base station at time k; γ is the attenuation coefficient. Let k be the response vector of the receiving antenna array at time k; θ is the azimuth angle of the receiving beam at the base station at time k; r,k Let be the elevation angle of the receiving beam at the base station at time k; (·) H Indicates conjugate transpose; Let k be the response vector of the transmitting antenna array at time k; Let θ be the azimuth angle of the transmitted beam at time k; t,k Let be the elevation angle of the transmitted beam at time k;

[0021] Based on the time-varying channel matrix of UAV-to-ground millimeter-wave communication, a base station receiving spectrum efficiency model is constructed:

[0022]

[0023] Among them, Rspe k (ω k ,f k ω is the spectral efficiency of the channel at time k; k f is the beam combination vector of the receiving antenna at time k; k Let be the beamforming vector of the transmitting antenna at time k; ρ is the average transmitted signal power. For the noise received at time k, The variance is the variance of Gaussian white noise with a mean of 0; |·| represents the absolute value.

[0024] Furthermore, the construction of the coordinate system to establish the relationship between the UAV attitude and the azimuth and elevation angles of the transmitted beam includes:

[0025] Construct a global coordinate system and a UAV coordinate system. The global coordinate system takes the first element of the receiving antenna array at the base station as its origin, the north direction of the base station as the x-axis, the east direction of the base station as the y-axis, and the z-axis perpendicular to the north and east directions of the base station. The UAV coordinate system takes the first element of the transmitting antenna array as its origin, the forward direction of the UAV as the x-axis, the rightward direction of the forward direction of the UAV as the y-axis, and the z-axis perpendicular to the forward direction of the UAV and its rightward direction.

[0026] The rotation angle of the UAV coordinate system's z-axis around the global coordinate system's z-axis is defined as the yaw angle φ. z The rotation angle of the y-axis of the UAV coordinate system around the y-axis of the global coordinate system is defined as the pitch angle φ. y The rotation angle of the UAV coordinate system's x-axis around the global coordinate system's x-axis is defined as the roll angle φ. x The azimuth and elevation angles of the UAV's transmitted beam are angles defined in the UAV's coordinate system. The UAV's three-dimensional spatial attitude is obtained by rotation relative to the global coordinate system. The UAV's coordinate system and the global coordinate system are transformed using the following formula:

[0027]

[0028] in, This is the transformation matrix between the global coordinate system and the UAV coordinate system;

[0029] Construct the UAV's transmit beam direction matrix in the UAV coordinate system at time k:

[0030]

[0031] in, Let be the transmission beam direction matrix of the UAV in the UAV coordinate system at time k; for The transpose of φ x,k φ is the roll angle at time k. y,k The pitch angle at time k; φ z,k Let k be the yaw angle; Let be the transmit beam direction matrix of the UAV in the global coordinate system at time k. Let k be the coordinates of the UAV in the global coordinate system at time k; The coordinates of the base station in the global coordinate system;

[0032] get:

[0033]

[0034] in, Let θ be the azimuth angle of the transmitted beam at time k; t,kLet be the elevation angle of the transmitted beam at time k; This represents the x-axis position of the base station in the UAV coordinate system after coordinate transformation. This represents the y-axis position of the base station in the UAV coordinate system after coordinate transformation. This represents the z-axis position of the base station in the UAV coordinate system after coordinate transformation.

[0035] Furthermore, the step of inputting the UAV trajectory prediction results, the extended Kalman filter estimation results of the UAV attitude, and the backpropagation estimation results into the receive spectral efficiency model to obtain the true spectral efficiency includes:

[0036] Calculate the beam reception azimuth and elevation angles at the base station in the global coordinate system using the following formulas:

[0037]

[0038] The base station side obtains the position of the UAV at time k in the global coordinate system through prediction. This refers to the beam receiving azimuth angle at the base station in the global coordinate system. The beam reception elevation angle at the base station in the global coordinate system; The coordinates of the base station in the global coordinate system;

[0039] The true spectral efficiency is calculated using the following formula:

[0040]

[0041] Among them, Rspe k Let k be the true spectral efficiency of the channel at time k. Let be the beam combination vector of the receiving antenna in the global coordinate system at time k; Let be the beamforming vector of the transmitting antenna in the global coordinate system at time k; ρ is the average transmitted signal power. For noise received at time k in the global coordinate system. The variance is to conform to Gaussian white noise with a mean of 0; |·| represents the absolute value; H pef,k Let k be the real space channel matrix at time k; and θ r,k,0 Let be the azimuth and elevation angles of the received beam at time k without positional error, respectively. and θ t,k,0 These are the azimuth and elevation angles of the transmitted beam at time k without attitude error, respectively. Let be the response vector of the receiving antenna array at time k without position error; Let be the transmit antenna array response vector at time k without attitude error.

[0042] Secondly, the present invention provides a beam tracking system based on UAV spatial positioning, comprising:

[0043] The model building module is used to build models of UAV-to-ground millimeter-wave communication systems and receiver spectrum efficiency models.

[0044] The coordinate system construction module is used to construct a coordinate system to establish the relationship between the UAV's attitude and the azimuth and elevation angles of the transmitted beam;

[0045] The attitude estimation module is used to determine the azimuth and elevation angles of the UAV's transmitted beam by performing extended Kalman filter estimation and backpropagation estimation on the UAV's attitude.

[0046] The trajectory prediction module is used to predict the trajectory of the UAV and obtain the UAV trajectory prediction result;

[0047] The spectrum efficiency calculation module is used to input the UAV trajectory prediction results, UAV attitude extended Kalman filter estimation results, and backpropagation estimation results into the receiving spectrum efficiency model to obtain the true spectrum efficiency, and to quantify the beam tracking effect with the true spectrum efficiency.

[0048] Thirdly, the present invention provides a computer device including a processor and a memory; wherein, when the processor executes a computer program stored in the memory, it implements the steps of the beam tracking method based on UAV spatial positioning described in the first aspect.

[0049] Fourthly, the present invention provides a computer-readable storage medium, characterized in that it is used to store a computer program; when the computer program is executed by a processor, it implements the steps of the beam tracking method based on UAV spatial positioning described in the first aspect.

[0050] This invention provides a beam tracking method and system based on UAV spatial positioning. The method combines a backpropagation neural network and an extended Kalman filter algorithm to estimate the UAV's attitude and predict its trajectory at a base station. By combining the estimation and prediction results, a spatial beam angle is obtained through coordinate transformation, forming a simulated beamforming vector to achieve beam tracking. This method can effectively achieve accurate beam tracking of UAVs in dynamic environments. Attached Figure Description

[0051] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0052] Figure 1A flowchart illustrating a beam tracking method based on UAV spatial positioning provided in an embodiment of the present invention;

[0053] Figure 2 This is a structural diagram of a beam tracking method based on UAV spatial positioning provided in an embodiment of the present invention;

[0054] Figure 3 A schematic diagram of spatial coordinates provided for an embodiment of the present invention;

[0055] Figure 4 A simulation diagram of attitude estimation provided for an embodiment of the present invention;

[0056] Figure 5 The UAV location prediction map provided in the embodiments of the present invention;

[0057] Figure 6 Simulation diagram of spectral efficiency versus antenna array size provided for embodiments of the present invention;

[0058] Figure 7 This is a structural diagram of a beam tracking system based on UAV spatial positioning provided in an embodiment of the present invention. Detailed Implementation

[0059] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0060] In one embodiment, such as Figure 1 As shown, this embodiment of the invention provides a beam tracking method based on UAV spatial positioning, including:

[0061] Step 101: Construct a model of the UAV-to-ground millimeter-wave communication system and a model of the receiving spectrum efficiency.

[0062] For example, construct a transmit antenna array response model:

[0063]

[0064] Among them, the drone's transmitting antenna is equipped with M t ×N t A uniform planar antenna array with individual antenna elements; This is the response vector of the transmitting antenna array; θ is the azimuth angle of the transmitted beam at the UAV location. t The elevation angle of the transmitted beam at the UAV location; the array element spacing of the transmitting antenna d1 = λ c / 2;λ c The carrier wavelength is 1 ≤ m t ≤M t ; 1≤n t ≤N t ;m t n is the number of rows in the uniform planar antenna array for transmitting antennas; t M represents the number of columns in the uniform planar antenna array for transmitting the antenna. In this embodiment, M... t =10; N t =16; λ c =0.005m.

[0065] Construct a receive antenna array response model:

[0066]

[0067] Among them, the base station receiving antenna is equipped with M r ×N r A uniform planar antenna array with individual antenna elements; This is the response vector of the receiving antenna array; θ is the azimuth angle of the receiving beam at the base station. r The elevation angle of the receiving beam at the base station; the array element spacing of the receiving antenna d2 = λ c / 2;1≤m r ≤M r ; 1≤n r ≤N r ;m r n is the number of rows in the uniform planar antenna array for receiving antennas; r M represents the number of columns in the uniform planar antenna array for receiving antennas. In this embodiment, M... r =16; N r =16.

[0068] Assuming both the base station and the drone employ analog beamforming systems with a single radio frequency chain, and since the drone's ground communication primarily relies on a line-of-sight channel for signal transmission, a time-varying channel matrix for drone-to-ground millimeter-wave communication is constructed based on the transmit and receive antenna array response models:

[0069]

[0070] Among them, H k Let be the time-varying channel matrix for UAV-to-ground millimeter-wave communication at time k; gain k D is the complex path gain at time k; k Let be the transmission distance between the drone and the base station at time k; γ is the attenuation coefficient. Let k be the response vector of the receiving antenna array at time k; θ is the azimuth angle of the receiving beam at the base station at time k; r,k Let be the elevation angle of the receiving beam at the base station at time k; (·) H Indicates conjugate transpose; Let k be the response vector of the transmitting antenna array at time k; Let θ be the azimuth angle of the transmitted beam at time k; t,k Let be the elevation angle of the transmitted beam at time k for the UAV. and θ t,k The value is calculated from the attitude estimation results; and θ r,k The value of is calculated from the location prediction result. In this embodiment of the invention, D k =40m, gain k Let be a time-varying complex number with a modulus less than 1, representing the attenuation and phase deflection of the time-varying channel, considering a free-space channel γ = 2.

[0071] Based on the time-varying channel matrix of UAV-to-ground millimeter-wave communication, a base station receiving spectrum efficiency model is constructed:

[0072]

[0073] Among them, Rspe k (ω k ,f k ω is the spectral efficiency of the channel at time k; k f is the beam combination vector of the receiving antenna at time k; k Let be the beamforming vector of the transmitting antenna at time k; ρ is the average transmitted signal power. For the noise received at time k, The variance is set to conform to Gaussian white noise with a mean of 0; |·| represents the absolute value. In the embodiments of this invention... ρ = 25.

[0074] During beam switching, this paper analyzes how the spatial positioning accuracy of a UAV affects the millimeter-wave channel. UAV spatial positioning includes determining its position and attitude. Position can be directly obtained at the UAV's location, while at the base station, its trajectory needs to be tracked using a prediction algorithm. In this embodiment, attitude is represented by the UAV's yaw angle, roll angle, and pitch angle in three-dimensional space. The azimuth and elevation angles of the transmit and receive beams in the channel matrix are determined based on the UAV's position and attitude.

[0075] like Figure 3 As shown, the ideal channel matrix is ​​defined as H pef That is, assuming no errors in the drone's position and attitude, at a certain altitude, communication between the drone and the base station consists solely of a line-of-sight wireless transmission channel, where H is at time k. pef for:

[0076]

[0077] in, and θ r,k,0 These are the azimuth and elevation angles of the received beam at time k without positional error, respectively. and θ t,k,0 These are the azimuth and elevation angles of the transmitted beam at time k without attitude error; H pef,k Let be the real space channel matrix at time k.

[0078] Since the base station is stationary and the surrounding environment remains largely unchanged, the receiving beam azimuth and elevation angles at the base station remain constant when only the UAV's attitude estimation error exists. The UAV's attitude estimation error only causes changes in the transmitting beam azimuth and elevation angles at the UAV's location. θ t,k,att =θ t,k,0 +Δθ t,k,att , and Δθ t,k,att Let represent the transmit beam direction angle error and elevation angle error caused by the UAV attitude estimation error at time k, respectively. The channel matrix at time k with attitude error is defined as:

[0079]

[0080] and θ t,k,att H represents the azimuth and elevation angles of the transmitted beam with errors calculated at time k due to attitude estimation errors. att,k The channel matrix is ​​used to design the transmit beam vector at the transmitter at time k.

[0081] Compared to attitude estimation errors, which only cause changes in the azimuth and elevation angles of the transmitted beam at the UAV, changes in the UAV's position cause changes in the azimuth and elevation angles of the received beam at the base station. Therefore, by adding position prediction errors to the above, the channel matrix with UAV positioning errors at time k is obtained as follows:

[0082]

[0083] and θ r,k,pos These are the azimuth and elevation angles of the received beam with error, calculated at time k due to the position prediction error.

[0084] Step 102: Construct a coordinate system to establish the relationship between the UAV attitude and the azimuth and elevation angles of the transmitted beam.

[0085] Establish a coordinate system to describe the position and attitude of the UAV. Calculate the azimuth and elevation angles of the transmitted and received beams through coordinate transformation. Construct a global coordinate system and a UAV coordinate system. The global coordinate system has its origin at the first element of the receiving antenna array at the base station, with the north direction of the base station as the x-axis, the east direction as the y-axis, and the z-axis perpendicular to the north and east directions. The UAV coordinate system has its origin at the first element of the transmitting antenna array, with the UAV's forward direction as the x-axis, the direction to the right of the forward direction as the y-axis, and the z-axis perpendicular to the forward direction and the direction to the right of the forward direction.

[0086] The rotation angle of the UAV coordinate system's z-axis around the global coordinate system's z-axis is defined as the yaw angle φ. z The rotation angle of the y-axis of the UAV coordinate system around the y-axis of the global coordinate system is defined as the pitch angle φ. y The rotation angle of the UAV coordinate system's x-axis around the global coordinate system's x-axis is defined as the roll angle φ. x The three-dimensional spatial attitude of the UAV at time k can be described as (φ) x,k ,φ y,k ,φ z,k The azimuth and elevation angles of the UAV's transmitted beam are angles defined in the UAV's coordinate system. The UAV's three-dimensional spatial attitude is obtained by rotation relative to the global coordinate system. The UAV's coordinate system and the global coordinate system are transformed using the following formula:

[0087]

[0088] in, This is the transformation matrix between the global coordinate system and the UAV coordinate system;

[0089] Construct the UAV's transmit beam direction matrix in the UAV coordinate system at time k:

[0090]

[0091] For example, suppose the base station is located at (0,0,10) in the global coordinate system, and the drone is located at... The direction of the UAV's transmission beam in the body coordinate system can be expressed as:

[0092]

[0093] in, Let be the transmission beam direction matrix of the UAV in the UAV coordinate system at time k; for The transpose of φ x,k φ is the roll angle at time k. y,k The pitch angle at time k; φ z,k Let k be the yaw angle; Let be the transmit beam direction matrix of the UAV in the global coordinate system at time k. Let k be the coordinates of the UAV in the global coordinate system at time k; The coordinates of the base station in the global coordinate system;

[0094] get:

[0095]

[0096] in, Let θ be the azimuth angle of the transmitted beam at time k; t,k Let be the elevation angle of the transmitted beam at time k; This represents the x-axis position of the base station in the UAV coordinate system after coordinate transformation. This represents the y-axis position of the base station in the UAV coordinate system after coordinate transformation. This represents the z-axis position of the base station in the UAV coordinate system after coordinate transformation.

[0097] Step 103: Determine the azimuth and elevation angles of the UAV's transmitted beam by performing extended Kalman filter estimation and backpropagation estimation on the UAV's attitude.

[0098] Extended Kalman Filter (EKF) attitude estimation and Backpropagation (BP) attitude estimation. For example... Figure 2 As shown, in UAV beam tracking, the azimuth and elevation angles of the transmitted beam are determined by the UAV's position and attitude. However, in reality, the UAV's three-dimensional spatial attitude cannot be directly obtained. The UAV's attitude needs to be calculated based on the sensor data carried by the UAV. This embodiment of the invention uses a combination of EKF and BP algorithms to calculate the UAV's attitude, reducing attitude estimation errors and thus improving beam tracking accuracy.

[0099] 1) Define the standard form of the state space:

[0100]

[0101] z k =h(x k )+v k (13)

[0102] in,

[0103]

[0104] Where f(·) and h(·) are the state transition model and measurement model, respectively, and x k z is a state variable. kLet w be the measured variable, Δt be the sampling interval, and w be the measured variable. k and v k For process noise and observation noise, w k ~(0,Q k ), v k ~(0,R k ), Q k Let R be the process noise covariance matrix. k To observe the noise covariance matrix.

[0105] 2) Establish a state transition model.

[0106] In this embodiment, data obtained from the three-axis accelerometer, three-axis gyroscope, and three-axis magnetometer mounted on the UAV are used to estimate the UAV's attitude. The accelerometer data at time k is denoted as acc. k =(acc x,k ,acc y,k ,acc z,k The gyroscope data at time k represents gyr. k =(gyr x,k gyr y,k gyr z,k The magnetometer reading at time k is expressed as mag = (mag x,k ,mag y,k ,mag z,k ).

[0107] The attitude of the UAV at time k (φ) x,k ,φ y,k ,φ z,k The quaternion form is represented as quat. k =(quat) 0,k jquat 1,k jquat 2,k jquat 3,k The conversion formula is:

[0108]

[0109] The gyroscope bias is represented as b. k =(b x,k ,b y,k ,b z,k If ), then the state variable can be represented as:

[0110] x k =(quat) 0,k ,quat 1,k ,quat 2,k ,quat 3,k ,b x,k ,b y,k ,bz,k (16)

[0111] The state transition function is:

[0112]

[0113] in

[0114]

[0115] n ω The model is zero-mean Gaussian white noise, where gyr is the actual measurement value of the gyr on the UAV, and the gyr = gyr - b is used to remove the gyr deviation.

[0116] 3) Establish a system measurement model.

[0117] The measured variable is an attitude quaternion, represented as:

[0118] z k =(quat) 0,k ,quat 1,k ,quat 2,k ,quat 3,k (19)

[0119] The outputs of the magnetometer and accelerometer can be directly converted into quaternions using the Wahba loss function or algebraic solutions. This embodiment employs an orthogonalization method, utilizing only the measurements from the accelerometer and magnetometer to obtain an orthogonal matrix, and then derives a unit quaternion. The rotation matrix consists of three column vectors i, j, and e, where i represents the x-axis of the UAV coordinate system, and j and e represent the y-axis and z-axis of the UAV coordinate system, respectively.

[0120] e = acc. (20)

[0121] i=mag-norm(acc)·(mag·norm(acc)). (twenty one)

[0122] norm(j)=norm(e)×norm(i). (twenty two)

[0123] Where norm is the vector normalization function, acc is the accelerometer measurement vector, and mag is the magnetometer measurement vector, the rotation matrix between the global coordinate system and the UAV coordinate system is obtained:

[0124]

[0125] Then, the unit pose quaternion is obtained from the rotation matrix:

[0126]

[0127] The measurement equation is:

[0128]

[0129] 4) Perform EKF pose estimation

[0130] Prior state estimation:

[0131]

[0132] Prior error covariance matrix:

[0133] P k|k-1 =Γ k-1 P k-1|k-1 Γ k-1 T +Q k-1 (27)

[0134] in

[0135]

[0136] E(·) represents the expected value, and the Kalman gain is:

[0137] K k =P k|k-1 U k T (U k P k|k-1 U k T +R k ) -1 (29)

[0138] in

[0139]

[0140] Posterior state estimation:

[0141]

[0142] The posterior error covariance matrix is:

[0143] P k|k =(IK k U k )P k|k-1 (32)

[0144] This includes the EKF attitude quaternion estimate, which is converted into Euler angles to obtain the UAV's three-dimensional spatial attitude (φ) at time k. x_ekf,k ,φ y_ekf,k ,φ z_ekf,k The conversion formula is as follows:

[0145]

[0146] UAV attitude estimation is a typical multivariate nonlinear vector regression problem. Backpropagation (BP) is an algorithm that combines forward propagation of signals and backward propagation of errors. The network structure can be adjusted according to the specific problem, and a model with generalization and fault tolerance can be generated through supervised learning.

[0147] In this embodiment of the invention, a four-layer BP neural network model is constructed, including one input layer, two hidden layers, and one output layer. To utilize the sequential features of pose changes, the BP input layer includes sensor data (gyr) from two time points. x,k-1 gyr y,k- gyr z,k-1 ,acc x,k-1 ,acc y,k-1 ,acc z,k-1 ,mag x,k-1 ,mag y,k-1 ,mag z,k-1 gyr x,k gyr y,k gyr z,k ,acc x,k ,acc y,k ,acc z,k ,mag x,k ,mag y,k ,mag z,k That is, the input layer has 18 neurons, and the output layer has (φ) x,k ,φ y,k ,φ z,k The network has 3 neurons in its core, and the hidden layers have 12 and 6 neurons respectively, with the activation function being logsig. This embodiment of the invention collects sensor data at 8000 time points, of which 6000 time points are used as training data for the BP network, and the remaining 2000 time points are used as test data. The resulting BP pose estimation result (φ) is obtained. x—bp,k ,φ y—bp,k ,φ z—bp,k ).

[0148] EKF-BP attitude estimation.

[0149] Since prior knowledge of sensor error characteristics cannot be accurately obtained in practice, the estimation results of the EKF algorithm have large errors. Therefore, the BP estimation results and the EKF estimation results are passed together through a BP network to improve the estimation accuracy and enhance robustness.

[0150] Similarly, a four-layer backpropagation (BP) neural network is constructed, consisting of one input layer, two hidden layers, and one output layer. The input layer contains the results of EKF and BP estimations (φ). x—ekf,k ,φy—ekf,k ,φ z—ekf,k ,φ x—bp,k ,φ y—bp,k ,φ z—bp,k It has 6 neuron nodes, and the output layer is (φ x,k ,φ y,k ,φ z,k The network has 3 neurons in its core, and 6 and 3 neurons in its hidden layers, respectively. The activation functions are logsig and purelin. This embodiment of the invention uses 8000 time-stamp data points, with 6000 time-stamps used as training data and the remaining 2000 time-stamps used as testing data. The trained network yields the final result of the UAV's 3D spatial pose estimation. Table 1 shows the attitude estimation results.

[0151] Will Substituting into equation (9), we obtain the azimuth and elevation angles of the transmit beam with attitude estimation error. Substituting into equation (6) again, we obtain the channel matrix with attitude estimation error.

[0152] Table 1 Attitude estimation error

[0153]

[0154] Step 104: Perform trajectory prediction on the UAV to obtain the UAV trajectory prediction result.

[0155] To address the issue of base stations being unable to obtain the real-time position of drones during autonomous movement, this invention employs a Gate Recurrent Unit (GRU) neural network to predict the drone's attitude, thereby achieving accurate beam tracking. It is assumed that the drone sends its historical position information to the base station at regular intervals as prediction input. GRU is an improved model based on recurrent neural networks, with a basic structure consisting of an input layer, hidden layers, and an output layer. The hidden layer contains memory units and gate structures. The memory units store past information, while the gate structures include update and reset gates to control the influence of historical and current information. GRU uses a current input... k The hidden state (hid) passed from the previous node k-1 (Including relevant information from previous nodes), combined with in k and hid k-1 GRU will obtain the output of the currently hidden node. k and the hidden state hid passed to the next node k The calculation of the memory unit update mechanism and gate structure control mechanism at time k is as follows:

[0156] Update gate formula:

[0157] up k =σ(W z in k +W z hid k-1 (34)

[0158] Reset door formula:

[0159] rn k =σ(W r in k +W r hid k-1 (35)

[0160] Calculate the current state based on the reset result. for:

[0161]

[0162] Calculate the hidden state hid k for:

[0163]

[0164] GRU output:

[0165] out k =σ(W o hid k (38)

[0166] In this embodiment of the invention, in k Using historical trajectory sequences as input, out k The predicted trajectory sequence is used as the output, W is the weight vector, σ is the sigma activation function, and tanh is the tanh activation function. The UAV position is defined as (x... n ,y n ,z n Training is performed separately for each dimension, such as the x-axis of the drone. n When tracking coordinates, the input is The output is Where S and S′ represent the number of historical trajectory time slots and the number of predicted time slots, respectively. This enables position tracking of the UAV at the base station, achieving beam tracking.

[0167] Taking the x-axis coordinate as an example, the input to the GRU network is a vector. The output is The y-axis and z-axis are predicted in the same way, with predictions made for 10 time points each time, meaning the drone sends its position to the base station every 10 time points. This invention uses continuous position data in the global coordinate system at 8000 time points, with 6000 time points used as training data and the remaining 2000 time points as test data. The position of the drone at time k in the global coordinate system is obtained through prediction.

[0168] Step 105: Input the UAV trajectory prediction results, UAV attitude extended Kalman filter estimation results, and backpropagation estimation results into the receiving spectral efficiency model to obtain the true spectral efficiency, and quantify the beam tracking effect with the true spectral efficiency.

[0169] The position of the UAV at time k in the global coordinate system was obtained through prediction at the base station. Calculate the beam reception azimuth and elevation angles at the base station in the global coordinate system using the following formulas:

[0170]

[0171] The base station side obtains the position of the UAV at time k in the global coordinate system through prediction. This refers to the beam receiving azimuth angle at the base station in the global coordinate system. The beam reception elevation angle at the base station in the global coordinate system; The coordinates of the base station in the global coordinate system.

[0172] The true spectral efficiency is calculated using the following formula:

[0173]

[0174] Among them, Rspe k Let k be the true spectral efficiency of the channel at time k. Let be the beam combination vector of the receiving antenna in the global coordinate system at time k; Let be the beamforming vector of the transmitting antenna in the global coordinate system at time k; ρ is the average transmitted signal power. For noise received at time k in the global coordinate system. The variance is to conform to Gaussian white noise with a mean of 0; |·| represents the absolute value; H pef,k Let k be the real space channel matrix at time k; and θ r,k,0 Let be the azimuth and elevation angles of the received beam at time k without positional error, respectively. and θ t,k,0 These are the azimuth and elevation angles of the transmitted beam at time k without attitude error, respectively. Let be the response vector of the receiving antenna array at time k without position error; Let be the transmit antenna array response vector at time k without attitude error. and The time-frequency efficiency is the highest.

[0175] The effect obtained in this embodiment can be achieved through… Figures 4 to 6 The specific data obtained in the simulation experiment will be further explained. We see that: Figure 4 The results of attitude estimation using the EKF-BP method and the EKF and BP algorithms alone show that the EKF fitting effect is better than the estimation results of the two methods alone. Figure 5 The results of predicting UAV trajectories using the GRU method and the Long-Short Term Memory (LSTM) neural network show that the GRU method has a significantly better fitting effect. Figure 6 To assess the spectral efficiency of beam tracking based on UAV spatial positioning, it can be seen that the EKF-BP algorithm for attitude estimation and the GRU algorithm for position prediction, i.e., the EKF-BP-GRU beam tracking method, results in the least spectral efficiency loss.

[0176] This invention combines a backpropagation neural network and an extended Kalman filter algorithm to estimate the attitude of a UAV, and uses a gated recursive unit at the base station to predict the UAV trajectory. By combining the estimation and prediction results, a spatial beam angle is obtained through coordinate transformation, forming a simulated beamforming vector to achieve beam tracking. This effectively enables precise beam tracking of UAVs in dynamic environments.

[0177] Based on the same inventive concept, this invention also provides a beam tracking system based on UAV spatial positioning. Since the principle of this system in solving the problem is similar to the above-mentioned beam tracking method based on UAV spatial positioning, the implementation of this system can refer to the implementation of the beam tracking method based on UAV spatial positioning, and the repeated parts will not be described again.

[0178] In another embodiment, the beam tracking system based on UAV spatial positioning provided by this invention, such as... Figure 7 As shown, it includes:

[0179] Model building module 10 is used to build a model of the UAV-to-ground millimeter-wave communication system and a model of the receiving spectrum efficiency.

[0180] The coordinate system construction module 20 is used to construct a coordinate system to establish the relationship between the UAV attitude and the azimuth and elevation angles of the transmitted beam.

[0181] Angle estimation module 30 is used to determine the azimuth and elevation angles of the UAV's transmitted beam by performing extended Kalman filter estimation and backpropagation estimation on the UAV's attitude.

[0182] The trajectory prediction module 40 is used to predict the trajectory of the UAV and obtain the UAV trajectory prediction result.

[0183] The spectrum efficiency calculation module 50 is used to input the UAV trajectory prediction results, the extended Kalman filter estimation results of the UAV attitude, and the backpropagation estimation results into the receiving spectrum efficiency model to obtain the true spectrum efficiency, and to quantify the beam tracking effect with the true spectrum efficiency.

[0184] For more detailed information on the working process of each of the above modules, please refer to the relevant content disclosed in the foregoing embodiments, which will not be repeated here.

[0185] In another embodiment, the present invention provides a computer device including a processor and a memory; wherein the processor executes a computer program stored in the memory to implement the steps of the above-described beam tracking method based on UAV spatial positioning.

[0186] For more detailed information on the above methods, please refer to the relevant content disclosed in the foregoing embodiments, which will not be repeated here.

[0187] In another embodiment, the present invention provides a computer-readable storage medium for storing a computer program; when the computer program is executed by a processor, it implements the steps of the above-described beam tracking method based on UAV spatial positioning.

[0188] For more detailed information on the above methods, please refer to the relevant content disclosed in the foregoing embodiments, which will not be repeated here.

[0189] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. The systems, devices, and storage media disclosed in the embodiments are described simply because they correspond to the methods disclosed in the embodiments; relevant details can be found in the method section.

[0190] Those skilled in the art will clearly understand that the techniques in the embodiments of the present invention can be implemented using software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solutions in the embodiments of the present invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or certain parts of the embodiments of the present invention.

[0191] The present invention has been described in detail above with reference to specific embodiments and exemplary examples; however, these descriptions should not be construed as limiting the present invention. Those skilled in the art will understand that various equivalent substitutions, modifications, or improvements can be made to the technical solutions and embodiments of the present invention without departing from the spirit and scope of the invention, and all such modifications and improvements fall within the scope of the present invention. The scope of protection of the present invention is defined by the appended claims.

Claims

1. A beam tracking method based on UAV spatial positioning, characterized in that, include: Construct a model of a UAV-to-ground millimeter-wave communication system and a model of its receiving spectrum efficiency; Construct a coordinate system to establish the relationship between the UAV's attitude and the azimuth and elevation angles of the transmitted beam; The azimuth and elevation angles of the UAV's transmitted beam are determined by performing extended Kalman filter estimation and backpropagation estimation on the UAV's attitude. Specifically, this determination involves: The data obtained from the three-axis accelerometer, three-axis gyroscope and three-axis magnetometer on the UAV are input into the extended Kalman filter algorithm to estimate the UAV's attitude, and the data obtained from the three-axis accelerometer, three-axis gyroscope and three-axis magnetometer are input into the first BP neural network to estimate the UAV's attitude. The UAV attitude estimated by the extended Kalman filter algorithm and the UAV attitude estimated by the first BP neural network are input into the second BP neural network to obtain the final UAV attitude estimation result. The final UAV attitude estimation result is substituted into the relationship between the UAV attitude and the azimuth and elevation angles of the transmitted beam to obtain the azimuth and elevation angles of the UAV transmitted beam. The trajectory prediction of the UAV is performed to obtain the UAV trajectory prediction result; the specific process of trajectory prediction of the UAV is as follows: inputting the historical trajectory sequence of the UAV into the GRU network to obtain the predicted position of the UAV in the global coordinate system; The UAV trajectory prediction results, UAV attitude extended Kalman filter estimation results, and backpropagation estimation results are input into the receiver spectral efficiency model to obtain the true spectral efficiency, and the beam tracking effect is quantified by the true spectral efficiency.

2. The beam tracking method based on UAV spatial positioning according to claim 1, characterized in that, The construction of the UAV-to-ground millimeter-wave communication system model and the receiving spectrum efficiency model includes: Construct a transmit antenna array response model: ； Among them, the drone transmitting antenna is equipped with M t × N t A uniform planar antenna array with individual antenna elements; This is the response vector of the transmitting antenna array; The azimuth angle of the transmitted beam at the UAV location; The elevation angle of the transmitted beam at the UAV location; the spacing between the array elements of the transmitting antenna. ; The carrier wavelength; 1 ≤ m t ≤ M t ;1≤ n t ≤ N t ; m t The number of rows in the uniform planar antenna array for transmitting antennas; n t The number of columns in the uniform planar antenna array for transmitting antennas; Construct a receive antenna array response model: ； Among them, the base station receiving antenna is equipped with M r × N r A uniform planar antenna array with individual antenna elements; This is the response vector of the receiving antenna array; The azimuth angle of the receiving beam at the base station; The elevation angle of the receiving beam at the base station; the spacing between the array elements of the receiving antenna. ;1≤ m r ≤ M r ;1≤ n r ≤ N r ; m r The number of rows in the uniform planar antenna array for receiving antennas; n r The number of columns in the uniform planar antenna array for receiving antennas; Based on the transmit antenna array response model and the receive antenna array response model, a time-varying channel matrix for UAV-to-ground millimeter-wave communication is constructed: ； in, for k Time-varying channel matrix for UAV-to-ground millimeter-wave communication; gain k for k Complex path gain at time step; D k for k The transmission distance between the drone and the base station at any given time; γ The attenuation coefficient; for k Receive the antenna array response vector at any time; for k The azimuth angle of the receiving beam at the base station at any given time; for k Elevation angle of the receiving beam at the base station at any given time; Indicates conjugate transpose; for k The transmit antenna array response vector at any given time; for k The azimuth angle of the transmission beam at the location of the drone at any given moment; for k The elevation angle of the transmission beam at the drone location at any given moment; Based on the time-varying channel matrix of UAV-to-ground millimeter-wave communication, a base station receiving spectrum efficiency model is constructed: ； in, for k The spectral efficiency of the channel at any given time; for k The beam combination vector of the receiving antenna at any given time; for k Beamforming vector of the transmitting antenna at any given time; ρ This represents the average transmitted power of the signal. ,for k Constantly receiving noise, The variance is the variance of Gaussian white noise with a mean of 0; |·| represents the absolute value.

3. The beam tracking method based on UAV spatial positioning according to claim 1, characterized in that, The construction of the coordinate system to establish the relationship between the UAV attitude and the azimuth and elevation angles of the transmitted beam includes: Construct a global coordinate system and a UAV coordinate system; the global coordinate system takes the first element of the receiving antenna array at the base station as its origin, and the north direction of the base station as its coordinate system. x The axis direction, with the base station's due east direction as... y Axial direction, z The axis is perpendicular to the north and east directions of the base station; the UAV coordinate system has its origin at the first element of the transmitting antenna array, and the UAV's forward direction is... x The axis direction, the direction directly to the right of the drone's forward movement is... y Axial direction, z The axis is perpendicular to the drone's forward direction and its direct right direction; The coordinate system of the drone z Axis around global coordinate system z The rotation angle of the shaft is defined as the yaw angle. UAV coordinate system y Axis around global coordinate system y The rotation angle of the axis is defined as the pitch angle. UAV coordinate system x Axis around global coordinate system x The rotation angle of the shaft is defined as the roll angle. The azimuth and elevation angles of the UAV's transmitted beam are angles defined in the UAV's coordinate system. The UAV's three-dimensional spatial attitude is obtained by rotation relative to the global coordinate system. The UAV's coordinate system and the global coordinate system are transformed using the following formula: ； in, This is the transformation matrix between the global coordinate system and the UAV coordinate system; Build k The transmit beam direction matrix of the UAV in the time-based UAV coordinate system: ； in, for k The transmit beam direction matrix of the UAV in the UAV coordinate system at any given time; for The transpose of the matrix; for k Roll angle at all times; for k Pitch angle at all times; for k Yaw angle at time; for k The transmit beam direction matrix of the UAV in the global coordinate system at any given time. , for k The coordinates of the drone in the global coordinate system at any time; The coordinates of the base station in the global coordinate system; get: ； in, for k The azimuth angle of the transmission beam at the location of the drone at any given moment; for k The elevation angle of the transmission beam at the drone location at any given moment; For the base station in the UAV coordinate system after coordinate transformation x Axial position; For the base station in the UAV coordinate system after coordinate transformation y Axial position; For the base station in the UAV coordinate system after coordinate transformation z Position in the axial direction.

4. The beam tracking method based on UAV spatial positioning according to claim 1, characterized in that, The process of inputting the UAV trajectory prediction results, the extended Kalman filter estimation results of the UAV attitude, and the backpropagation estimation results into the receiver spectral efficiency model to obtain the true spectral efficiency includes: Calculate the beam reception azimuth and elevation angles at the base station in the global coordinate system using the following formulas: ； Among them, the base station side obtains the global coordinate system through prediction. k Location of drones ; This refers to the beam receiving azimuth angle at the base station in the global coordinate system. The beam reception elevation angle at the base station in the global coordinate system; , , () represents the coordinates of the base station in the global coordinate system; The true spectral efficiency is calculated using the following formula: ； in, for k The true spectral efficiency of the channel at any given time; for k The beam combination vector of the receiving antenna in the global coordinate system at any given time; for k The beamforming vector of the transmitting antenna in the global coordinate system at any given time; ρ This represents the average transmitted power of the signal. In the global coordinate system k Constantly receiving noise, The variance is to conform to Gaussian white noise with a mean of 0; |·| represents the absolute value; ; for k Real-time spatial channel matrix; and They are respectively k The azimuth and elevation angles of the receiving beam with no positional error at any given time. and They are respectively k The azimuth and elevation angles of the transmitted beam are always free of attitude errors; for k The response vector of the receiving antenna array with no positional error at any given time; for k The transmit antenna array response vector with no attitude error at any given time.

5. A beam tracking system based on UAV spatial positioning, characterized in that, include: The model building module is used to build models of UAV-to-ground millimeter-wave communication systems and receiver spectrum efficiency models. The coordinate system construction module is used to construct a coordinate system to establish the relationship between the UAV's attitude and the azimuth and elevation angles of the transmitted beam; The attitude estimation module is used to determine the azimuth and elevation angles of the UAV's transmitted beam by performing extended Kalman filter estimation and backpropagation estimation on the UAV's attitude; specifically, determining the azimuth and elevation angles of the UAV's transmitted beam by performing extended Kalman filter estimation and backpropagation estimation on the UAV's attitude involves: The data obtained from the three-axis accelerometer, three-axis gyroscope and three-axis magnetometer on the UAV are input into the extended Kalman filter algorithm to estimate the UAV's attitude, and the data obtained from the three-axis accelerometer, three-axis gyroscope and three-axis magnetometer are input into the first BP neural network to estimate the UAV's attitude. The UAV attitude estimated by the extended Kalman filter algorithm and the UAV attitude estimated by the first BP neural network are input into the second BP neural network to obtain the final UAV attitude estimation result. The final UAV attitude estimation result is substituted into the relationship between the UAV attitude and the azimuth and elevation angles of the transmitted beam to obtain the azimuth and elevation angles of the UAV transmitted beam. The trajectory prediction module is used to predict the trajectory of the UAV and obtain the UAV trajectory prediction result; the trajectory prediction of the UAV specifically involves: inputting the historical trajectory sequence of the UAV into the GRU network to obtain the predicted position of the UAV in the global coordinate system; The spectrum efficiency calculation module is used to input the UAV trajectory prediction results, UAV attitude extended Kalman filter estimation results, and backpropagation estimation results into the receiving spectrum efficiency model to obtain the true spectrum efficiency, and to quantify the beam tracking effect with the true spectrum efficiency.

6. A computer device, characterized in that, It includes a processor and a memory; wherein, when the processor executes the computer program stored in the memory, it implements the steps of the beam tracking method based on UAV spatial positioning as described in any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, Used to store computer programs; when the computer programs are executed by a processor, they implement the steps of the beam tracking method based on UAV spatial positioning as described in any one of claims 1-4.

Images